CN106094027A - A kind of vertical seismic profiling (VSP) VSP pre-drilling pressure forecasting method and system - Google Patents

Abstract

The invention discloses a kind of vertical seismic profiling (VSP) VSP pre-drilling pressure forecasting method and system, including: VSP earthquake record is carried out wave field separation, it is thus achieved that upgoing wave earthquake record and down going wave earthquake record.Down going wave earthquake record is carried out deconvolution process, extracts optimum deconvolution operator.Corridor stack earthquake record is obtained after being applied to optimum deconvolution operator on upgoing wave earthquake record process.According to VSP speed, Well-Log Acoustic Velocity and surface seismic speed comprehensive analysis processing, layering section sets up inverting initial velocity model.According to the inverting initial velocity model set up, corridor stack earthquake record is carried out Inversion Calculation, obtain VSP seismic impedance.The velocity of longitudinal wave of rock is calculated according to the VSP seismic impedance obtained.Velocity o P wave value formation pressure according to the rock calculated.Pass through the solution of the present invention, it is possible to comprehensive utilization VSP data, well-log information and surface-seismic data, improve stability and the precision of velocity inversion.

Description

A kind of vertical seismic profiling (VSP) VSP pre-drilling pressure forecasting method and system

Technical field

The present invention relates to field of seismic exploration, particularly relate to a kind of vertical seismic profiling (VSP) VSP pre-drilling pressure forecasting method and System.

Background technology

Along with the development of vertical seismic profiling (VSP) VSP technology is with ripe, its range of application in Exploration of Oil And Gas More extensive, therefore, the process for VSP data needs to reach the requirement of higher precision.VSP explores owing to entering in well Row excites or receives, and the seismic data collected has higher resolution compared with surface seismic.By itself and surface seismic, well logging money Material combines, and carries out formation velocity inverting, can obtain the formation velocity of higher precision.

In drilling process, it is often necessary to the stratum in drill bit front is carried out anticipation, superpressure layer is predicted, prevent from boring The generation of well accident.Formation velocity can increase along with the degree of depth and increase, and when there is superpressure layer in stratum, can cause formation velocity Bust.Therefore, calculate the strata pressure in drill bit front, thus carrying out prediction before drilling is that an existing theory significance has again actual answering By the technology being worth.Can not only stressor layer computational methods galore, provide theoretical foundation for subsurface geological structure research, also simultaneously Drilling well risk can be avoided, for oil-gas exploration and exploitation service.

At present conventional pre-drilling pressure forecasting method has variation, as based on logging method, surface seismic method etc., often The method of kind respectively has superiority with not enough.Forecasting Methodology based on surface-seismic data can obtain large range of predicting the outcome, but Being that surface-seismic data is relatively low in deep layer resolution, the prediction of result precision of gained is relatively low；Prediction before drilling based on well-log information Method is typically statistically to analyze, the well-log information explored in this block of predicted well place due between well and well Difference, it was predicted that result is only used for qualitative analysis.

Present at present velocity inversion and prediction before drilling technique study, subject matter is: 1, how to comprehensively utilize all kinds of surveying Spy data sets up inverting initial velocity model；2, how seismic wavelet estimation more accurately is obtained according to VSP down going wave data；3、 How to use more effective inversion method, improve stability and the inversion accuracy of refutation process.

Summary of the invention

In order to solve the problems referred to above, the present invention proposes a kind of vertical seismic profiling (VSP) VSP pre-drilling pressure forecasting method and is System, it is possible to comprehensive utilization VSP data, well-log information and surface-seismic data, improves stability and the precision of velocity inversion.

In order to achieve the above object, the present invention proposes a kind of vertical seismic profiling (VSP) VSP pre-drilling pressure forecasting method, the party Method includes:

The VSP earthquake record collected is carried out wave field separation, it is thus achieved that upgoing wave earthquake record and down going wave earthquake record.

Down going wave earthquake record is carried out deconvolution process, extracts optimum deconvolution operator.

Optimum deconvolution operator is applied on upgoing wave earthquake record, after process, obtains corridor stack earthquake record.

Inverting is set up the fastest according to the VSP speed obtained in advance, Well-Log Acoustic Velocity and surface seismic speed layering section Degree model.

According to the inverting initial velocity model set up, the inversion algorithm preset is used to carry out anti-corridor stack earthquake record Drill calculating, obtain VSP seismic impedance.

The velocity of longitudinal wave of rock is calculated according to the VSP seismic impedance obtained.

According to the velocity of longitudinal wave of the rock calculated, the first algorithm preset is used to calculate value formation pressure.

Alternatively, down going wave earthquake record is carried out deconvolution process, extracts optimum deconvolution operator and include:

Extract deconvolution operator after down going wave earthquake record is carried out deconvolution process, and obtain expected wavelet；Adjust anti- Convolution operator, makes to be less than or equal to default first through the down going wave earthquake record of deconvolution process with the error of expected wavelet Error threshold；And the deconvolution operator after adjusting is as optimum deconvolution operator.

Alternatively, optimum deconvolution operator is applied on upgoing wave earthquake record, after process, obtains corridor stack earthquake Record includes:

Use optimum deconvolution operator upgoing wave earthquake record is carried out deconvolution process, make through deconvolution process upper Shake record in row rolling land is less than or equal to the second default error threshold with the error of expected wavelet.

The upgoing wave earthquake record processed through deconvolution carries out static time shift process with corridor stack, it is thus achieved that corridor is folded Add earthquake record.

Alternatively,

VSP speed refers to: down going wave earthquake record carries out the interval velocity of velocity analysis acquisition.

Well-Log Acoustic Velocity refers to: the interval transit time obtained after drilling well section is by acoustic logging obtains after changing Interval velocity.

Surface seismic speed refers to: surface seismic record carries out the seismic velocity spectrum of velocity analysis acquisition.

Alternatively,

Inverting is set up the fastest according to the VSP speed obtained in advance, Well-Log Acoustic Velocity and surface seismic speed layering section Degree model includes:

There is the interval of Well-Log Acoustic Velocity, the rate curve of Well-Log Acoustic Velocity is being removed exceptional value and to do filtering flat The curve obtained after sliding process is as inverting initial velocity model.

In the rate curve of Well-Log Acoustic Velocity, there is no the interval of Well-Log Acoustic Velocity, using the speed of VSP speed Curve makes up the rate curve of the interval not having Well-Log Acoustic Velocity, to carrying out Well-Log Acoustic Velocity that rate curve makes up Rate curve removes exceptional value, and does the curve obtained after filtering processes as inverting initial velocity model；Or, entirely When portion's interval does not all have Well-Log Acoustic Velocity, obtain after the rate curve of VSP speed being removed exceptional value and doing filtering process The curve obtained is as inverting initial velocity model.

In the rate curve of Well-Log Acoustic Velocity, there is no Well-Log Acoustic Velocity and the interval of VSP speed, by VSP ground Shake record mates with surface seismic record, extracts record in the seismic velocity spectrum recording corresponding interval with VSP earthquake The seismic velocity of drilling well position, using the seismic velocity extracted to make up does not has Well-Log Acoustic Velocity and the interval of VSP speed Rate curve, be filtered the rate curve carrying out the Well-Log Acoustic Velocity that rate curve makes up after smoothing processing obtaining Curve as inverting initial velocity model；Or, all there is no Well-Log Acoustic Velocity and the interval of VSP speed at whole intervals, Being mated with surface seismic record by VSP earthquake record, extraction records with VSP earthquake in the seismic velocity spectrum of corresponding interval The seismic velocity of the drilling well position of record, it is thus achieved that the rate curve of the seismic velocity extracted, enters the rate curve of acquisition The curve that filtering of going obtains after processing is as inverting initial velocity model.

Alternatively,

The inversion algorithm preset is inversion algorithm based on bayesian theory.

According to the inverting initial velocity model set up, the inversion algorithm preset is used to carry out anti-corridor stack earthquake record Drill calculating, obtain VSP seismic impedance and include:

Simulating the observation data to obtain in advance as dependent variable, the model parameter preset in inverting initial velocity model is The linear matrix of independent variable.

Utilize first equation first position in inverting initial velocity model that corridor stack earthquake record is launched, obtain The real seismic record of first position that must be relevant to model parameter.

Using the difference of the real seismic record of first position and corridor stack earthquake record as the disturbance observing data Amount.

After linear matrix is carried out pretreatment, combine with linear matrix acquisition the first relation by the disturbance quantity of observation data Formula.

First relational expression is introduced Bayesian formula, it is thus achieved that the Posterior distrbutionp relational expression of observation data.

According to the disturbance quantity of model parameter in each iterative process in Posterior distrbutionp relational expression acquisition Inversion Calculation.

Combine with inverting initial velocity model acquisition final speed curve by the disturbance quantity of model parameter.

According to final speed curve and the well density curve acquisition VSP seismic impedance that obtains in advance.

Alternatively, include according to the velocity of longitudinal wave of the VSP seismic impedance calculating rock obtained: by VSP seismic impedance Substitute into the velocity of longitudinal wave calculating rock in the second algorithm preset.

The method also includes: VSP seismic impedance is substituting into the velocity of longitudinal wave calculating rock in the second algorithm preset Before, according to the well logging sound wave in the VSP earthquake record being obtained ahead of time and the density data in well density curve to the second algorithm It is fitted, simulates the parameter value in the second algorithm；The parameter value simulated is substituted into the second algorithm, and matching will be included Second algorithm of the parameter value gone out is as the equation of the velocity of longitudinal wave calculating rock.

Alternatively,

First algorithm is the Fillippone formula improved.

Second algorithm is Gardner formula.

First equation is Taylor's formula.

In order to achieve the above object, the invention allows for a kind of vertical seismic profiling (VSP) VSP pre-drilling pressure forecasting system, should System includes: separation module, the first processing module, the second processing module, model building module, inverting module, the first calculating mould Block and the second computing module.

Separation module, for carrying out wave field separation to the VSP earthquake record collected, it is thus achieved that upgoing wave earthquake record with under Row rolling land shake record.

First processing module, for down going wave earthquake record carries out deconvolution process, extracts optimum deconvolution operator.

Second processing module, for optimum deconvolution operator is applied to upgoing wave earthquake record, obtains after process Corridor summed seismogram.

Model building module, for according to the VSP speed obtained in advance, Well-Log Acoustic Velocity and the layering of surface seismic speed Duan Jianli inverting initial velocity model.

Inverting module, for according to the inverting initial velocity model set up, presetting the employing of corridor stack earthquake record Inversion algorithm carries out Inversion Calculation, obtains VSP seismic impedance.

First computing module, for calculating the velocity of longitudinal wave of rock according to the VSP seismic impedance obtained.

Second computing module, for the velocity of longitudinal wave according to the rock calculated, uses the first algorithm preset to calculate ground Lamination force value.

Alternatively, the first processing module carries out deconvolution process to down going wave earthquake record, extracts optimum deconvolution operator Including:

Extract deconvolution operator after down going wave earthquake record is carried out deconvolution process, and obtain expected wavelet；And adjust Deconvolution operator, makes the error of the down going wave earthquake record and the expected wavelet that process through deconvolution less than or equal to the preset One error threshold；And the deconvolution operator after adjusting is as optimum deconvolution operator.

Alternatively, optimum deconvolution operator is applied on upgoing wave earthquake record by the second processing module, obtains after process Corridor stack earthquake record includes:

Use optimum deconvolution operator upgoing wave earthquake record is carried out deconvolution process, make through deconvolution process upper Shake record in row rolling land is less than or equal to the second default error threshold with the error of expected wavelet.

The upgoing wave earthquake record processed through deconvolution carries out static time shift process with corridor stack, it is thus achieved that corridor is folded Add earthquake record.

Alternatively,

VSP speed refers to: down going wave earthquake record carries out the interval velocity of velocity analysis acquisition.

Well-Log Acoustic Velocity refers to: the interval transit time obtained after drilling well section is by acoustic logging obtains after changing Interval velocity.

Surface seismic speed refers to: surface seismic record carries out the seismic velocity spectrum of velocity analysis acquisition.

Alternatively,

Model building module is built according to the VSP speed obtained in advance, Well-Log Acoustic Velocity and surface seismic speed layering section Vertical inverting initial velocity model includes:

There is the interval of Well-Log Acoustic Velocity, the rate curve of Well-Log Acoustic Velocity is being removed exceptional value and to do filtering flat The curve obtained after sliding process is as inverting initial velocity model.

In the rate curve of Well-Log Acoustic Velocity, there is no the interval of Well-Log Acoustic Velocity, using the speed of VSP speed Curve makes up the rate curve of the interval not having Well-Log Acoustic Velocity, to carrying out Well-Log Acoustic Velocity that rate curve makes up Rate curve removes exceptional value, and does the curve obtained after filtering processes as inverting initial velocity model；Or, entirely When portion's interval does not all have Well-Log Acoustic Velocity, obtain after the rate curve of VSP speed being removed exceptional value and doing filtering process The curve obtained is as inverting initial velocity model.

In the rate curve of Well-Log Acoustic Velocity, there is no Well-Log Acoustic Velocity and the interval of VSP speed, by VSP ground Shake record mates with surface seismic record, extracts record in the seismic velocity spectrum recording corresponding interval with VSP earthquake The seismic velocity of drilling well position, using the seismic velocity extracted to make up does not has Well-Log Acoustic Velocity and the interval of VSP speed Rate curve, be filtered the rate curve carrying out the Well-Log Acoustic Velocity that rate curve makes up after smoothing processing obtaining Curve as inverting initial velocity model；Or, all there is no Well-Log Acoustic Velocity and the interval of VSP speed at whole intervals, Being mated with surface seismic record by VSP earthquake record, extraction records with VSP earthquake in the seismic velocity spectrum of corresponding interval The seismic velocity of the drilling well position of record, it is thus achieved that the rate curve of the seismic velocity extracted, enters the rate curve of acquisition The curve that filtering of going obtains after processing is as inverting initial velocity model.

Alternatively,

The inversion algorithm preset is inversion algorithm based on bayesian theory.

Corridor stack earthquake record, according to the inverting initial velocity model set up, is used the inverting preset to calculate by inverting module Method carries out Inversion Calculation, obtains VSP seismic impedance and includes:

Simulating the observation data to obtain in advance as dependent variable, the model parameter preset in inverting initial velocity model is The linear matrix of independent variable.

Utilize first equation first position in inverting initial velocity model that corridor stack earthquake record is launched, obtain The real seismic record of first position that must be relevant to model parameter.

Using the difference of the real seismic record of first position and corridor stack earthquake record as the disturbance observing data Amount.

After linear matrix is carried out pretreatment, combine with linear matrix acquisition the first relation by the disturbance quantity of observation data Formula.

First relational expression is introduced Bayesian formula, it is thus achieved that the Posterior distrbutionp relational expression of observation data.

According to the disturbance quantity of model parameter in each iterative process in Posterior distrbutionp relational expression acquisition Inversion Calculation.

Combine with inverting initial velocity model acquisition final speed curve by the disturbance quantity of model parameter.

According to final speed curve and the well density curve acquisition VSP seismic impedance that obtains in advance.

Alternatively, the velocity of longitudinal wave that the first computing module calculates rock according to the VSP seismic impedance obtained includes: will VSP seismic impedance substitutes into the velocity of longitudinal wave calculating rock in the second algorithm preset.

The method also includes: VSP seismic impedance is substituting into the velocity of longitudinal wave calculating rock in the second algorithm preset Before, according to the well logging sound wave in the VSP earthquake record being obtained ahead of time and the density data in well density curve to the second algorithm It is fitted, simulates the parameter value in the second algorithm；The parameter value simulated is substituted into the second algorithm, and matching will be included Second algorithm of the parameter value gone out is as the equation of the velocity of longitudinal wave calculating rock.

Alternatively,

First algorithm is the Fillippone formula improved.

Second algorithm is Gardner formula.

First equation is Taylor's formula.

Compared with prior art, the present invention includes: the VSP earthquake record collected is carried out wave field separation, it is thus achieved that up Rolling land shake record and down going wave earthquake record.Down going wave earthquake record is carried out deconvolution process, extracts optimum deconvolution operator. Optimum deconvolution operator is applied on upgoing wave earthquake record, after process, obtains corridor stack earthquake record.This process energy Enough reduce the seismic wavelet impact on inversion result in inverting.According to the VSP speed, Well-Log Acoustic Velocity and the ground that obtain in advance Seismic velocity is comprehensively analyzed, and layering section sets up inverting initial velocity model.According to the inverting initial velocity model set up, right Corridor stack earthquake record uses the inversion algorithm preset to carry out Inversion Calculation, obtains VSP seismic impedance.According to obtain VSP seismic impedance calculates the velocity of longitudinal wave of rock.According to the velocity of longitudinal wave of the rock calculated, use the first algorithm preset Calculate value formation pressure.Pass through the solution of the present invention, it is possible to comprehensive utilization VSP data, well-log information and surface-seismic data, carry The stability of inverting and precision at high speed.

Accompanying drawing explanation

Illustrating the accompanying drawing in the embodiment of the present invention below, the accompanying drawing in embodiment is for the present invention is entered one Step understands, is used for explaining the present invention, is not intended that limiting the scope of the invention together with description.

Fig. 1 is vertical seismic profiling (VSP) VSP pre-drilling pressure forecasting method flow diagram；

Fig. 2 is vertical seismic profiling (VSP) VSP pre-drilling pressure forecasting system block diagram.

Detailed description of the invention

For the ease of the understanding of those skilled in the art, the invention will be further described below in conjunction with the accompanying drawings, not Can be used for limiting the scope of the invention.

In drilling process, it is often necessary to the stratum in drill bit front is carried out anticipation, superpressure layer is predicted, prevent from boring The generation of well accident.Formation velocity can increase along with the degree of depth and increase, and when there is superpressure layer in stratum, can cause formation velocity Bust.And downhole receiving to VSP data can the most directly reflect the stratum in the range of the certain depth of drill bit front undoubtedly Velocity variations situation.Therefore, utilize VSP data to carry out velocity inversion, and calculate the strata pressure in drill bit front accordingly, thus enter Row prediction before drilling is the technology that an existing theory significance has again actual application value.Formation pressure calculation side can not only be enriched Method, provides theoretical foundation for subsurface geological structure research, can also avoid drilling well risk simultaneously, for oil-gas exploration and exploitation service.

The most conventional pre-drilling pressure forecasting method has variation, as based on logging method, surface seismic method and VSP Methods etc., every kind of method has superiority with not enough.Forecasting Methodology based on surface-seismic data can obtain large range of pre- Surveying result, but surface-seismic data is relatively low in deep layer resolution, the prediction of result precision of gained is relatively low；Based on well-log information Prediction before drilling method is typically statistically to analyze, the well-log information explored in this block of predicted well place due to well And the difference between well, it was predicted that result is only used for qualitative analysis；And VSP data is relevant to drill bit earth layer in front practical situation Property the highest exploration data, compared with well-log information, it can find out the information not boring stratum, has again simultaneously and provides than surface seismic Expect higher resolution, if VSP data can be carried out high-precision inverting to obtain the speed of drill bit earth layer in front, degree of depth letter Breath, predicting the outcome of obtaining will more conform to the practical situation on stratum.

It is a nonlinear inverse problem that geological data carries out velocity inversion, and its multi-solution situation is serious, but VSP number According to having its innate advantage, time and depth transfer relation and seismic wavelet accurately can be obtained from log data and VSP data and estimate Meter.Existing inverting method for solving has a lot, is broadly divided into classical inverting method for solving and inverting side based on statistics idea Method.Use inverting based on statistics idea can better profit from the prior information of all different scales, as log, Shake normal-moveout spectrum, has very great help for improving inversion accuracy.

Seismic data is utilized to carry out prediction of formation pressure, mainly by the low speed feature of superpressure layer.Conventional pressure Computing formula includes relying on the equation of normal compaction trend line and being independent of formula two class of normal compaction trend line: Qian Zhezhu Equivalent depth computing method of formula to be included (Reynold, 1974), Easton method (1976) and Stone method (1983), the latter is then Including Fillippone method (1978,1982), Liu Zhenfa (1990) and Martinez method (1986) etc..Rely on compaction trend line The pressure prediction of computational methods is the most accurate, the stratum that especially buried depth is bigger, but the error of compaction curve by accumulation to pressure During power is evaluated, the foundation of compaction curve simultaneously can only the experience of domain of dependence, the most practical for wild-cat area；And it is independent of normal The empirical equation of pressure trend line should use more convenient, and in, the formation pressure calculation of the shallow degree of depth there is certain standard Really property, is more suitable for the superpressure layer prediction of wild-cat area.

It is an object of the invention to provide a kind of high accuracy speed comprehensively utilizing VSP data, well-log information, surface-seismic data Degree inverting and pre-drilling pressure forecasting method.

The present invention is to have studied seismic velocity inversion method based on bayesian theory, based on VSP down going wave record Propose on the basis of Method for Seismic Wavelet Estimation, the feasibility that velocity inversion result is applied in formation pressure calculation and effectiveness Come.

In order to achieve the above object, the present invention proposes a kind of vertical seismic profiling (VSP) VSP pre-drilling pressure forecasting method, such as figure Shown in 1, the method includes:

S101, the VSP earthquake record collected is carried out wave field separation, it is thus achieved that upgoing wave earthquake record and down going wave earthquake Record.

In embodiments of the present invention, vertical seismic profiling (VSP) technology refers to: arrange epicenter excitation seismic wave on earth's surface, in well Cymoscope is installed and receives seismic wave, i.e. observe one-dimensional artificial field in vertical direction, then to being observed the data obtained through school Just, superposition, filtering etc. process, obtain vertical seismic profiling (VSP).

The VSP earthquake record collected, is mainly made up of upgoing wave and down going wave, as in figure 2 it is shown, down going wave wave field is main By direct wave, transmitted wave, descending repeatedly wave component, upgoing wave wave field is mainly involved multiple reflection by primary event and forms.Depend on According to the difference of apparent velocity between uplink and downlink ripple Yu polarization direction, VSP earthquake record can be separated into upgoing wave earthquake record and under Row rolling land shake record.

S102, down going wave earthquake record is carried out deconvolution process, extract optimum deconvolution operator.

Deconvolution (deconvolution) is also known as inverse filtering (inverse filter) or deconvolution.Eliminate the most previous Plant the processing method of filter action.It it is the processing procedure being improved geological data vertical resolution by compression basic wavelet.? Ideally, deconvolution energy compact wavelet length MULTIPLE ATTENUATION, on tunnel, finally only retain underlying reflection coefficient.

When inverse filter makes its impulse response and signal convolution exactly, the filtering being added on signal before some can be eliminated and make With.Such as seismic wave is propagated in stratum, and stratum can be regarded as the wave filter with certain character.Therefore, it can pass through These filter actions are removed by deconvolution, recover the shape of excitation signal approx, to improve resolution capability.Inverse filter kind A lot, the inverse filter designed for eliminating reverberation, ghosting, multiple reflections and other interference is mainly included.

Alternatively, down going wave earthquake record is carried out deconvolution process, extracts optimum deconvolution operator and include:

Extract deconvolution operator after down going wave earthquake record is carried out deconvolution process, and obtain expected wavelet；Adjust anti- Convolution operator, makes to be less than or equal to default first through the down going wave earthquake record of deconvolution process with the error of expected wavelet Error threshold；And the deconvolution operator after adjusting is as optimum deconvolution operator.

Through the step for process after, in refutation process later, can directly use expected wavelet to carry out inverting.This One operation can avoid the error additionally introduced during extracting wavelet.

S103, optimum deconvolution operator is applied on upgoing wave earthquake record, after process, obtains corridor stack earthquake note Record.

Alternatively, optimum deconvolution operator is applied on upgoing wave earthquake record, after process, obtains corridor stack earthquake Record includes:

S1031, employing optimum deconvolution operator carry out deconvolution process to upgoing wave earthquake record, make at deconvolution The upgoing wave earthquake record of reason and the error of expected wavelet are less than or equal to the second error threshold preset.

S1032, the upgoing wave earthquake record processed through deconvolution carried out static time shift and corridor stack process, it is thus achieved that Corridor stack earthquake record.

In embodiments of the present invention, static time shift refers to: the VSP for zero-offset horizontal interface observes, it is assumed that in well Cymoscope all receives the echo t0 that direct wave t1, upgoing wave t2, down going wave (secondary) t3 and ground receiver arrive, then have t2+ T1=t3-t1=0, if each for upgoing wave road is all added first arrival time, is then equivalent to that cymoscope is put into wellhead ground and receives Launch the echo at interface, then upgoing wave will be come into line, the process adding first arrival time by its two-way time from the end table to interface It is static time shift.

Corridor stack refers to: on the section of static time shift (or static correction and come into line), from the oblique lineups of preliminary wave to many On one band (passage) of subwave termination line (oblique line), for once ripple, and excised many subwaves, primary wave homophase Axle is added together, forms single seismic channel, and this work is called corridor stack.

Corridor stack record can be approximately zero-offset VSP earthquake record, is the reflection record of self excitation and self receiving in well. On this basis, corridor stack record can be assumed to be the convolution of seismic wavelet and stratum reflection coefficient, can use convolution mould Type carries out forward simulation.Therefore, it can solve formation velocity by inversion algorithm.

VSP speed, Well-Log Acoustic Velocity and surface seismic speed layering section that S104, basis obtain in advance are set up at the beginning of inverting Beginning rate pattern.

In embodiments of the present invention, in seismic inversion, due to the earthquake record disappearance to low-frequency information itself, at the beginning of foundation During beginning rate pattern, low frequency trend is the most important.Here, low frequency refers to stratum or the Changing Pattern of some feature of stratum.This Method combines VSP speed, Well-Log Acoustic Velocity and surface seismic velocity analysis to the substantially layer position of prediction interval position and velocity variations Trend carries out anticipation, thus sets up low frequency trend accurately.In embodiments of the present invention, inverting initial velocity model refers to inverting Applied in reflection interval velocity change approximate trend rate curve.

Alternatively,

VSP speed refers to: down going wave earthquake record carries out the interval velocity of velocity analysis acquisition.

Well-Log Acoustic Velocity refers to: the interval transit time obtained after drilling well section is by acoustic logging obtains after changing Interval velocity.

Surface seismic speed refers to: surface seismic record carries out the seismic velocity spectrum of velocity analysis acquisition.

Alternatively, comprehensively analyze according to the VSP speed obtained in advance, Well-Log Acoustic Velocity and surface seismic speed, layering Duan Jianli inverting initial velocity model includes:

There is the interval of Well-Log Acoustic Velocity, the rate curve of Well-Log Acoustic Velocity is being removed exceptional value and to do filtering flat The curve obtained after sliding process is as inverting initial velocity model.

In the rate curve of Well-Log Acoustic Velocity, there is no the interval of Well-Log Acoustic Velocity, using the speed of VSP speed Curve, according to the rate curve of the interval of existing Well-Log Acoustic Velocity, the speed making up the interval not having Well-Log Acoustic Velocity is bent Line, removes exceptional value to the rate curve carrying out the Well-Log Acoustic Velocity that rate curve makes up, and does after filtering processes The curve obtained is as inverting initial velocity model；Or, when whole intervals all do not have Well-Log Acoustic Velocity, by VSP speed Rate curve remove exceptional value and do filtering process after obtain curve as inverting initial velocity model.

In the rate curve of Well-Log Acoustic Velocity, there is no Well-Log Acoustic Velocity and the interval of VSP speed, by VSP ground Shake record mates with surface seismic record, extracts record in the seismic velocity spectrum recording corresponding interval with VSP earthquake The seismic velocity of drilling well position, uses the seismic velocity extracted more, according to the speed of the interval of existing Well-Log Acoustic Velocity Curve, mends the rate curve of the interval not having Well-Log Acoustic Velocity and VSP speed, to carrying out the well logging sound that rate curve makes up The curve that the rate curve of wave velocity obtains after being filtered smoothing processing is as inverting initial velocity model；Or, all Interval does not all have Well-Log Acoustic Velocity and the interval of VSP speed, is mated with surface seismic record by VSP earthquake record, takes out Take the seismic velocity recording the drilling well position recorded in the seismic velocity spectrum of corresponding interval with VSP earthquake, it is thus achieved that extract The rate curve of seismic velocity, the curve carrying out the rate curve of acquisition obtaining after filtering processes is as at the beginning of inverting Beginning rate pattern.

Low frequency trend inverting initial velocity model the most accurately can be set up by such scheme.

S105, according to set up inverting initial velocity model, to corridor stack earthquake record use preset inversion algorithm Carry out Inversion Calculation, obtain VSP seismic impedance.

Alternatively, the inversion algorithm preset is inversion algorithm based on bayesian theory.

In embodiments of the present invention, owing to seismic velocity inversion exists multi-solution and the problem of refutation process instability, This, use inversion algorithm based on bayesian theory to calculate.

Alternatively, according to the inverting initial velocity model set up, corridor stack earthquake record use the inverting preset calculate Method carries out Inversion Calculation, obtains VSP seismic impedance and includes:

S1051, simulate the observation data to obtain in advance as dependent variable, the model preset in inverting initial velocity model Parameter is the linear matrix of independent variable.

In embodiments of the present invention, the inversion problem of the natural impedance of described corridor stack earthquake record can be approximated by we For linearization problem, use the form of descriptor matrix to represent here, simulate the observation data to obtain in advance as dependent variable, instead Drill the linear matrix that model parameter is independent variable preset in initial velocity model, specific as follows shown:

D=G (m)+n (5.1)

Wherein, d represents the observation data obtained in advance, and G is operation matrix, and m is default model parameter, and n is noise.

S1052, utilize first equation first position in inverting initial velocity model to corridor stack earthquake record Launch, it is thus achieved that the real seismic record of the first position relevant to model parameter.

Alternatively, the first equation is Taylor's formula.

In embodiments of the present invention, according to generalization inverting thought, utilize Taylor's formula by D_iIn initial velocity model phase The S answered_iPlace launches:

$D_i=S_i+\underset{k=0}{\overset{n-1}{\Σ}}{\mathrm{\Δm}}_k\frac{\∂S_i}{\∂m_k}+\frac{1}{2}{\mathrm{\Δm}}_k^2\left(\underset{k=0}{\overset{n-1}{\Σ}}{\mathrm{\Δm}}_k\frac{{\∂}^2{S}_i}{\∂m_k\∂m_j}\right)+......,---\left(5.2\right)$

Wherein, D_iFor real seismic record；S_iFor the synthetic seismogram that inversion speed initial model is corresponding, i.e. inverting speed The corridor stack earthquake record that degree initial model is corresponding；m_kFor the model parameter at k；Δm_kModel parameter disturbance for respective point Amount.The i in the most above-mentioned formula of primary importance in such scheme.

S1053, using the difference of the real seismic record of first position and corridor stack earthquake record as observation data Disturbance quantity.

In embodiments of the present invention, orderWherein W is that wavelet constitutes square Battle array, r is reflection coefficient.

S1054, linear matrix is carried out pretreatment after, combine with linear matrix acquisition the by the disturbance quantity of observation data One relational expression.

In embodiments of the present invention, (5.1) formula is omitted high-order term, in the case of muting, can be changed into:

$\mathrm{\Δ}d-G*\mathrm{\Δ}m=\frac{1}{2}{\mathrm{\Δm}}_k^2\left(\underset{k=0}{\overset{n-1}{\Σ}}{\mathrm{\Δm}}_k\frac{{\∂}^2{S}_i}{\∂m_k\∂m_j}\right).---\left(5.3\right)$

The first relational expression in relational expression (5.3) i.e. the present invention program.

S1055, by first relational expression introduce Bayesian formula, it is thus achieved that observation data Posterior distrbutionp relational expression.

In embodiments of the present invention, relational expression (5.3) is introduced Bayesian formula:

$P\left(m\right|d,I)=\frac{P\left(d\right|m,I)P\left(m\right|I)}{P\left(d\right|I)},---\left(5.4\right)$

Due to observation data it is known that then P (d | I) be a constant, and

$\begin{array}{c}P\left(m\right|I)=P(m=m_{l-1}+{\mathrm{\Δm}}_l|I)\\ =P({\mathrm{\Δm}}_l=m_{l-1}-m_l|I)\\ =P\left({\mathrm{\Δm}}_l\right|I)\end{array},---\left(5.5\right)$

Above formula can be reduced to:

P (m | d, I)=1/K P (d | m, I) P (Δ m | I). (5.6)

When there is no noise, and P (d | m, I) with (5.3) formula, there is identical probability distribution, it is assumed that (5.3) formula left end meets The Gauss distribution of zero-mean, and variance is equally distributed, then have:

$P\left(d\right|m,I)=\frac{1}{{\left(\mathrm{\σ}\sqrt{2\mathrm{\π}}\right)}^N}\×\exp \[-\frac{\underset{n=1}{\overset{N}{\Σ}}{({\mathrm{\Δd}}_n-G*{\mathrm{\Δm}}_n)}^2}{2{\mathrm{\σ}}^2}\],---\left(5.7\right)$

Wherein, n is sampled point sequence number, and N is total number of sample points；σ is error in data, i.e. the standard deviation of Δ d-G* Δ m.Equally The disturbance quantity of hypothesized model parameter also meets the Gauss distribution of zero-mean, then have:

$P\left(m\right|I)=\frac{1}{\sqrt{\det |C_{\mathrm{\Δ}m}{|}^3}2{\mathrm{\π}}^{\frac{3}{2}}}\×\exp \[-\frac{-{\mathrm{\Δm}}^T\mathrm{\Δ}m}{2C_{\mathrm{\Δ}m}}\],---\left(5.8\right)$

Wherein C_ΔmVariance for model parameter disturbance quantity.

Can be at the Posterior distrbutionp under observation data determine then:

$\begin{array}{c}P\left(m\right|d,I)=P\left(d\right|m,I)P\left(m\right|I)\\ =\frac{1}{{\left(\mathrm{\σ}\sqrt{2\mathrm{\π}}\right)}^N}\×\exp \[-\frac{\underset{n=1}{\overset{N}{\Σ}}{({\mathrm{\Δd}}_n-G*{\mathrm{\Δm}}_n)}^2}{2{\mathrm{\σ}}^2}\]\×\frac{1}{\sqrt{\det |C_{\mathrm{\Δ}m}{|}^3}2{\mathrm{\π}}^{\frac{3}{2}}}\×\exp \[-\frac{-{\mathrm{\Δm}}^T\mathrm{\Δ}m}{2C_{\mathrm{\Δ}m}}\]\end{array}.---\left(5.9\right)$

S1056, obtain in Inversion Calculation every time the disturbance quantity of model parameter in iterative process according to Posterior distrbutionp relational expression.

In embodiments of the present invention, above formula (5.9) is taken the logarithm, and omits and solve unrelated parameter, can obtain:

$O\left(m\right|d,I)=-\frac{1}{2{\mathrm{\σ}}^2}\underset{n=1}{\overset{N}{\Σ}}{({\mathrm{\Δd}}_n-G*{\mathrm{\Δm}}_n)}^2-\frac{-{\mathrm{\Δm}}^T\mathrm{\Δ}m}{2C_{\mathrm{\Δ}m}},---\left(5.10\right)$

To above formula (5.10) derivation, and derivative zero setting is obtained:

$\mathrm{\Δ}m=-{\[G^{T}G+\frac{{\mathrm{\σ}}^2}{C_{\mathrm{\Δ}m}}I\]}^{-1}{G}^T\mathrm{\Δ}d.---\left(5.11\right)$

The most just can obtain in inversion algorithm the disturbance quantity of model parameter in each iterative process, i.e. obtain inversion algorithm In each iterative process disturbance quantity to rate pattern.

S1057, the disturbance quantity of model parameter is combined with inverting initial velocity model acquisition final speed curve.

S1058, according to final speed curve and the well density curve acquisition VSP seismic impedance that obtains in advance.

Natural impedance (wave impedance) is seismic wave when propagating in media as well, act on pressure on certain area with The ratio of the unit interval interior particle flow (i.e. area takes advantage of Particle Vibration Velocity) perpendicular through this area, has resistant implication, It is referred to as natural impedance, its numerical value product equal to Media density ρ and velocity of wave V.

The VSP seismic impedance that S106, foundation obtain calculates the velocity of longitudinal wave of rock.

Alternatively, include according to the velocity of longitudinal wave of the VSP seismic impedance calculating rock obtained: by VSP seismic impedance Substitute into the velocity of longitudinal wave calculating rock in the second algorithm preset.

The method also includes: VSP seismic impedance is substituting into the velocity of longitudinal wave calculating rock in the second algorithm preset Before, according to the well logging sound wave in the VSP earthquake record being obtained ahead of time and the density data in well density curve to the second algorithm It is fitted, simulates the parameter value in the second algorithm；The parameter value simulated is substituted into the second algorithm, and matching will be included Second algorithm of the parameter value gone out is as the equation of the velocity of longitudinal wave calculating rock.

Alternatively, the second algorithm is Gardner formula.

In embodiments of the present invention, Gardner formula is the empirical equation describing demonstration speed with density relationship,

$\mathrm{\ρ}=0.31V_P^{0.25},---\left(6.1\right)$

Wherein, ρ is rock density, and unit is g/cm³, V_PFor rock velocity of longitudinal wave.This formula is one under different lithology Average formula, if directly using empirical equation parameter, result of calculation can there is bigger error for specific region, therefore, The present invention is previously according to existing well logging sound wave, density data, to formulaIn parameter a and b be fitted.

The velocity of longitudinal wave of the rock that S107, basis calculate, uses the first algorithm preset to calculate value formation pressure.

Alternatively, the first algorithm is the Fillippone formula improved.

This method uses the Fillippone formula of the improvement proposed by Liu Zhen, according to formation interval velocity, the most above-mentioned rock The velocity o P wave value formation pressure of stone, and with the addition of pressure correction factor γ on this basis:

$P_f=\mathrm{\γ}\·\frac{ln(v_i/v_{max})}{ln(v_{\min }/v_{max})}{P}_{ov},---\left(7.1\right)$

Wherein, v_iFor formation interval velocity；v_max、v_minIt is respectively porosity close to zero-sum rigidity close to stratum when zero Speed, the former is similar to matrix velocity, and the latter is similar to space fluid velocity；P_ovFor overlying formation pressure.This formula is also base In the empirical equation that actual measurement data matching obtains.When there being known depth point pressure measured value, can be according to the existing degree of depth Deviation between point pressure measured value and empirical equation result of calculation calculates pressure correction factor γ, enters this formula result of calculation Row correction, thus obtain value formation pressure more accurately.

In order to achieve the above object, the invention allows for a kind of vertical seismic profiling (VSP) VSP pre-drilling pressure forecasting system 01, As in figure 2 it is shown, this system includes: separation module the 02, first processing module the 03, second processing module 04, model building module 05, Inverting module the 06, first computing module 07 and the second computing module 08.

Separation module 02, for carrying out wave field separation to the VSP earthquake record collected, it is thus achieved that upgoing wave earthquake record with Down going wave earthquake record.

First processing module 03, for down going wave earthquake record carries out deconvolution process, extracts optimum deconvolution operator.

Second processing module 04, for optimum deconvolution operator is applied to upgoing wave earthquake record, obtains after process Corridor stack earthquake record.

Model building module 05, for building according to the VSP speed, Well-Log Acoustic Velocity and the surface seismic speed that obtain in advance Vertical inverting initial velocity model.

Inverting module 06, for according to the inverting initial velocity model set up, uses corridor stack earthquake record and presets Inversion algorithm carry out Inversion Calculation, obtain VSP seismic impedance.

First computing module 07, for calculating the velocity of longitudinal wave of rock according to the VSP seismic impedance obtained.

Second computing module 08, for the velocity of longitudinal wave according to the rock calculated, uses the first algorithm preset to calculate Value formation pressure.

Alternatively, the first processing module 03 carries out deconvolution process to down going wave earthquake record, extracts optimal deconvolution and calculates Attached bag includes:

Extract deconvolution operator after down going wave earthquake record is carried out deconvolution process, and obtain expected wavelet；And adjust Deconvolution operator, makes the error of the down going wave earthquake record and the expected wavelet that process through deconvolution less than or equal to the preset One error threshold；And the deconvolution operator after adjusting is as optimum deconvolution operator.

Alternatively, optimum deconvolution operator is applied on upgoing wave earthquake record by the second processing module 04, obtains after process Obtain corridor stack earthquake record to include:

Use optimum deconvolution operator upgoing wave earthquake record is carried out deconvolution process, make through deconvolution process upper Shake record in row rolling land is less than or equal to the second default error threshold with the error of expected wavelet.

The upgoing wave earthquake record processed through deconvolution carries out static time shift process with corridor stack, it is thus achieved that corridor is folded Add earthquake record.

Alternatively,

VSP speed refers to: down going wave earthquake record carries out the interval velocity of velocity analysis acquisition.

Well-Log Acoustic Velocity refers to: the interval transit time obtained after drilling well section is by acoustic logging obtains after changing Interval velocity.

Surface seismic speed refers to: surface seismic record carries out the seismic velocity spectrum of velocity analysis acquisition.

Alternatively,

Model building module 05 is according to the VSP speed obtained in advance, Well-Log Acoustic Velocity and surface seismic speed total score Analysis processes, and layering section is set up inverting initial velocity model and included:

There is the interval of Well-Log Acoustic Velocity, the rate curve of Well-Log Acoustic Velocity is being removed exceptional value and to do filtering flat The curve obtained after sliding process is as inverting initial velocity model.

In the rate curve of Well-Log Acoustic Velocity, there is no the interval of Well-Log Acoustic Velocity, using the speed of VSP speed Curve, according to the rate curve of the interval of existing Well-Log Acoustic Velocity, the speed making up the interval not having Well-Log Acoustic Velocity is bent Line, removes exceptional value to the rate curve carrying out the Well-Log Acoustic Velocity that rate curve makes up, and does after filtering processes The curve obtained is as inverting initial velocity model；Or, when whole intervals all do not have Well-Log Acoustic Velocity, by VSP speed Rate curve remove exceptional value and do filtering process after obtain curve as inverting initial velocity model.

In the rate curve of Well-Log Acoustic Velocity, there is no Well-Log Acoustic Velocity and the interval of VSP speed, by VSP ground Shake record mates with surface seismic record, extracts record in the seismic velocity spectrum recording corresponding interval with VSP earthquake The seismic velocity of drilling well position, uses the seismic velocity extracted more, according to the speed of the interval of existing Well-Log Acoustic Velocity Curve, mends the rate curve of the interval not having Well-Log Acoustic Velocity and VSP speed, to carrying out the well logging sound that rate curve makes up The curve that the rate curve of wave velocity obtains after being filtered smoothing processing is as inverting initial velocity model；Or, all Interval does not all have Well-Log Acoustic Velocity and the interval of VSP speed, is mated with surface seismic record by VSP earthquake record, takes out Take the seismic velocity recording the drilling well position recorded in the seismic velocity spectrum of corresponding interval with VSP earthquake, it is thus achieved that extract The rate curve of seismic velocity, the curve carrying out the rate curve of acquisition obtaining after filtering processes is as at the beginning of inverting Beginning rate pattern.

Alternatively,

The inversion algorithm preset is inversion algorithm based on bayesian theory.

Inverting module 06, according to the inverting initial velocity model set up, uses the inverting preset to corridor stack earthquake record Algorithm carries out Inversion Calculation, obtains VSP seismic impedance and includes:

Simulating the observation data to obtain in advance as dependent variable, the model parameter preset in inverting initial velocity model is The linear matrix of independent variable.

Utilize first equation first position in inverting initial velocity model that corridor stack earthquake record is launched, obtain The real seismic record of first position that must be relevant to model parameter.

Using the difference of the real seismic record of first position and corridor stack earthquake record as the disturbance observing data Amount.

After linear matrix is carried out pretreatment, combine with linear matrix acquisition the first relation by the disturbance quantity of observation data Formula.

First relational expression is introduced Bayesian formula, it is thus achieved that the Posterior distrbutionp relational expression of observation data.

According to the disturbance quantity of model parameter in each iterative process in Posterior distrbutionp relational expression acquisition Inversion Calculation.

Combine with inverting initial velocity model acquisition final speed curve by the disturbance quantity of model parameter.

According to final speed curve and the well density curve acquisition VSP seismic impedance that obtains in advance.

Alternatively, the velocity of longitudinal wave that the first computing module 07 calculates rock according to the VSP seismic impedance obtained includes: will VSP seismic impedance substitutes into the velocity of longitudinal wave calculating rock in the second algorithm preset.

The method also includes: VSP seismic impedance is substituting into the velocity of longitudinal wave calculating rock in the second algorithm preset Before, according to the well logging sound wave in the VSP earthquake record being obtained ahead of time and the density data in well density curve to the second algorithm It is fitted, simulates the parameter value in the second algorithm；The parameter value simulated is substituted into the second algorithm, and matching will be included Second algorithm of the parameter value gone out is as the equation of the velocity of longitudinal wave calculating rock.

Alternatively,

First algorithm is the Fillippone formula improved.

Second algorithm is Gardner formula.

First equation is Taylor's formula.

Compared with prior art, the present invention includes: the VSP earthquake record collected is carried out wave field separation, it is thus achieved that up Rolling land shake record and down going wave earthquake record.Down going wave earthquake record is carried out deconvolution process, extracts optimum deconvolution operator. Optimum deconvolution operator is applied on upgoing wave earthquake record, after process, obtains corridor stack earthquake record.Reduce and extract ground The extra error that shake Wavelet processing introduces.Comprehensive according to the VSP speed obtained in advance, Well-Log Acoustic Velocity and surface seismic speed Analyzing and processing, layering section sets up inverting initial velocity model.According to the inverting initial velocity model set up, to corridor stack earthquake Record uses the inversion algorithm preset to carry out Inversion Calculation, obtains VSP seismic impedance.According to the VSP seismic impedance obtained Calculate the velocity of longitudinal wave of rock.According to the velocity of longitudinal wave of the rock calculated, the first algorithm preset is used to calculate strata pressure Value.Pass through the solution of the present invention, it is possible to comprehensive utilization VSP data, well-log information and surface-seismic data, improve velocity inversion Stability and precision.

Understand it should be noted that embodiment described above is for only for ease of those skilled in the art, and It is not used in and limits the scope of the invention, on the premise of without departing from the inventive concept of the present invention, those skilled in the art couple Any obvious replacement that the present invention is made and improvement etc. are all within protection scope of the present invention.

Claims (16)

1. a vertical seismic profiling (VSP) VSP pre-drilling pressure forecasting method, it is characterised in that described method includes:

The VSP earthquake record collected is carried out wave field separation, it is thus achieved that upgoing wave earthquake record and down going wave earthquake record；

Described down going wave earthquake record is carried out deconvolution process, extracts optimum deconvolution operator；

Described optimum deconvolution operator is applied on described upgoing wave earthquake record, after process, obtains corridor stack earthquake note Record；

Inverting initial velocity mould is set up according to the VSP speed obtained in advance, Well-Log Acoustic Velocity and surface seismic speed layering section Type；

According to the described inverting initial velocity model set up, the inversion algorithm preset is used to enter described corridor stack earthquake record Row Inversion Calculation, obtains VSP seismic impedance；

The velocity of longitudinal wave of rock is calculated according to the described VSP seismic impedance obtained；

According to the velocity of longitudinal wave of the described rock calculated, the first algorithm preset is used to calculate value formation pressure.

2. VSP pre-drilling pressure forecasting method as claimed in claim 1, it is characterised in that

Described described down going wave earthquake record is carried out deconvolution process, extracts optimum deconvolution operator and include:

Extract deconvolution operator after described down going wave earthquake record is carried out deconvolution process, and obtain expected wavelet；Adjust institute State deconvolution operator, make the error of described down going wave earthquake record and described expected wavelet that processes through deconvolution less than or etc. In the first default error threshold；And the described deconvolution operator after adjusting is as described optimum deconvolution operator.

3. VSP pre-drilling pressure forecasting method as claimed in claim 2, it is characterised in that described by described optimal deconvolution calculation Son is applied on described upgoing wave earthquake record, obtains corridor stack earthquake record and include after process:

Use described optimum deconvolution operator that described upgoing wave earthquake record is carried out deconvolution process, make to process through deconvolution The error of described upgoing wave earthquake record and described expected wavelet less than or equal to the second error threshold preset；

The described upgoing wave earthquake record processed through deconvolution carries out static time shift and corridor stack process, it is thus achieved that described in walk Corridor summed seismogram.

4. VSP pre-drilling pressure forecasting method as claimed in claim 1, it is characterised in that

Described VSP speed refers to: described down going wave earthquake record carries out the interval velocity of velocity analysis acquisition；

Described Well-Log Acoustic Velocity refers to: the interval transit time obtained after drilling well section is by acoustic logging obtains after changing Interval velocity；

Described surface seismic speed refers to: surface seismic record carries out the seismic velocity spectrum of velocity analysis acquisition.

5. VSP pre-drilling pressure forecasting method as claimed in claim 4, it is characterised in that

It is the fastest that VSP speed, Well-Log Acoustic Velocity and the surface seismic speed layering section that described basis obtains in advance sets up inverting Degree model includes:

There is the interval of described Well-Log Acoustic Velocity, the rate curve of described Well-Log Acoustic Velocity is being removed exceptional value and filters The curve obtained after ripple smoothing processing is as described inverting initial velocity model；

In the rate curve of described Well-Log Acoustic Velocity, there is no the interval of described Well-Log Acoustic Velocity, using described VSP speed The rate curve of degree makes up the rate curve of the interval not having described Well-Log Acoustic Velocity, to carrying out the institute that rate curve makes up The rate curve stating Well-Log Acoustic Velocity removes exceptional value, and does the curve obtained after filtering processes as at the beginning of described inverting Beginning rate pattern；Or, when whole intervals all do not have described Well-Log Acoustic Velocity, the rate curve of described VSP speed is gone Fall exceptional value and do the curve obtained after filtering processes as described inverting initial velocity model；

In the rate curve of described Well-Log Acoustic Velocity, there is no described Well-Log Acoustic Velocity and the interval of described VSP speed, Being mated with described surface seismic record by described VSP earthquake record, extraction records corresponding interval with described VSP earthquake The seismic velocity of the drilling well position of record in described seismic velocity spectrum, using the described seismic velocity extracted to make up does not has institute State the rate curve of the interval of Well-Log Acoustic Velocity and described VSP speed, to carrying out the described well logging sound that rate curve makes up The curve that the rate curve of wave velocity obtains after being filtered smoothing processing is as described inverting initial velocity model；Or, All interval does not all have described Well-Log Acoustic Velocity and the interval of described VSP speed, by described VSP earthquake record and described ground Earthquake record mates, and extraction records the drilling well recorded in the described seismic velocity spectrum of corresponding interval with described VSP earthquake The seismic velocity of position, it is thus achieved that the rate curve of the described seismic velocity extracted, is carried out the described rate curve obtained Cross the curve obtained after filtering processes as described inverting initial velocity model.

6. VSP pre-drilling pressure forecasting method as claimed in claim 1, it is characterised in that

Described default inversion algorithm is inversion algorithm based on bayesian theory；

Described according to the described inverting initial velocity model set up, described corridor stack earthquake record use the inverting preset calculate Method carries out Inversion Calculation, obtains VSP seismic impedance and includes:

Simulating the observation data to obtain in advance as dependent variable, the model parameter preset in described inverting initial velocity model is The linear matrix of independent variable；

Utilize first equation first position in described inverting initial velocity model to described corridor stack earthquake record exhibition Open, it is thus achieved that the real seismic record of the described first position relevant to described model parameter；

Using the difference of the described real seismic record of described first position and described corridor stack earthquake record as described sight Survey the disturbance quantity of data；

After described linear matrix is carried out pretreatment, combine with described linear matrix acquisition by the disturbance quantity of described observation data First relational expression；

Described first relational expression is introduced Bayesian formula, it is thus achieved that the Posterior distrbutionp relational expression of described observation data；

The disturbance quantity of model parameter described in each iterative process in Inversion Calculation is obtained according to described Posterior distrbutionp relational expression；

Combine with described inverting initial velocity model acquisition final speed curve by the disturbance quantity of described model parameter；

According to VSP seismic impedance described in described final speed curve and the well density curve acquisition that obtains in advance.

7. VSP pre-drilling pressure forecasting method as claimed in claim 6, it is characterised in that described according to the described VSP ground obtained The velocity of longitudinal wave of seismic wave impedance computation rock includes: is substituted in the second algorithm preset by described VSP seismic impedance and calculates The velocity of longitudinal wave of described rock；

Described method also includes: calculate the vertical of described rock in the second algorithm preset being substituted into by described VSP seismic impedance Before wave velocity, according to the well logging sound wave in the described VSP earthquake record being obtained ahead of time and the density in described well density curve Described second algorithm is fitted by data, simulates the parameter value in described second algorithm；The described parameter value that will simulate Substitute into described second algorithm, and using include the described parameter value simulated the second algorithm as the compressional wave calculating described rock The equation of speed.

8. VSP pre-drilling pressure forecasting method as claimed in claim 7, it is characterised in that

Described first algorithm is the Fillippone formula improved；

Described second algorithm is Gardner formula；

Described first equation is Taylor's formula.

9. a vertical seismic profiling (VSP) VSP pre-drilling pressure forecasting system, it is characterised in that described system includes: separation module, One processing module, the second processing module, model building module, inverting module, the first computing module and the second computing module；

Described separation module, for carrying out wave field separation to the VSP earthquake record collected, it is thus achieved that upgoing wave earthquake record with under Row rolling land shake record；

Described first processing module, for described down going wave earthquake record is carried out deconvolution process, extracts optimal deconvolution and calculates Son；

Described second processing module, for described optimum deconvolution operator is applied to described upgoing wave earthquake record, processes Rear acquisition corridor stack earthquake record；

Described model building module, for according to the VSP speed obtained in advance, Well-Log Acoustic Velocity and the layering of surface seismic speed Duan Jianli inverting initial velocity model；

Described inverting module, for according to the described inverting initial velocity model set up, adopting described corridor stack earthquake record Carry out Inversion Calculation with default inversion algorithm, obtain VSP seismic impedance；

Described first computing module, for calculating the velocity of longitudinal wave of rock according to the described VSP seismic impedance obtained；

Described second computing module, for the velocity of longitudinal wave according to the described rock calculated, uses the first algorithm meter preset Calculate value formation pressure.

10. VSP pre-drilling pressure forecasting system as claimed in claim 9, it is characterised in that

Described first processing module carries out deconvolution process to described down going wave earthquake record, extracts optimum deconvolution operator bag Include:

Extract deconvolution operator after described down going wave earthquake record is carried out deconvolution process, and obtain expected wavelet；Adjust institute State deconvolution operator, make the error of described down going wave earthquake record and described expected wavelet that processes through deconvolution less than or etc. In the first default error threshold；And the described deconvolution operator after adjusting is as described optimum deconvolution operator.

11. VSP pre-drilling pressure forecasting systems as claimed in claim 10, it is characterised in that described second processing module is by institute State optimum deconvolution operator to be applied on described upgoing wave earthquake record, obtain corridor stack earthquake record after process and include:

Use described optimum deconvolution operator that described upgoing wave earthquake record is carried out deconvolution process, make to process through deconvolution The error of described upgoing wave earthquake record and described expected wavelet less than or equal to the second error threshold preset；

The described upgoing wave earthquake record processed through deconvolution carries out static time shift and corridor stack process, it is thus achieved that described in walk Corridor summed seismogram.

12. VSP pre-drilling pressure forecasting systems as claimed in claim 9, it is characterised in that

Described VSP speed refers to: described down going wave earthquake record carries out the interval velocity of velocity analysis acquisition；

Described Well-Log Acoustic Velocity refers to: the interval transit time obtained after drilling well section is by acoustic logging obtains after changing Interval velocity；

Described surface seismic speed refers to: surface seismic record carries out the seismic velocity spectrum of velocity analysis acquisition.

13. VSP pre-drilling pressure forecasting systems as claimed in claim 12, it is characterised in that

Described model building module is built according to the VSP speed obtained in advance, Well-Log Acoustic Velocity and surface seismic speed layering section Vertical inverting initial velocity model includes:

There is the interval of described Well-Log Acoustic Velocity, the rate curve of described Well-Log Acoustic Velocity is being removed exceptional value and filters The curve obtained after ripple smoothing processing is as described inverting initial velocity model；

In the rate curve of described Well-Log Acoustic Velocity, there is no the interval of described Well-Log Acoustic Velocity, using described VSP speed The rate curve of degree makes up the rate curve of the interval not having described Well-Log Acoustic Velocity, to carrying out the institute that rate curve makes up The rate curve stating Well-Log Acoustic Velocity removes exceptional value, and does the curve obtained after filtering processes as at the beginning of described inverting Beginning rate pattern；Or, when whole intervals all do not have described Well-Log Acoustic Velocity, the rate curve of described VSP speed is gone Fall exceptional value and do the curve obtained after filtering processes as described inverting initial velocity model；

In the rate curve of described Well-Log Acoustic Velocity, there is no described Well-Log Acoustic Velocity and the interval of described VSP speed, Being mated with described surface seismic record by described VSP earthquake record, extraction records corresponding interval with described VSP earthquake The seismic velocity of the drilling well position of record in described seismic velocity spectrum, using the described seismic velocity extracted to make up does not has institute State the rate curve of the interval of Well-Log Acoustic Velocity and described VSP speed, to carrying out the described well logging sound that rate curve makes up The curve that the rate curve of wave velocity obtains after being filtered smoothing processing is as described inverting initial velocity model；Or, All interval does not all have described Well-Log Acoustic Velocity and the interval of described VSP speed, by described VSP earthquake record and described ground Earthquake record mates, and extraction records the drilling well recorded in the described seismic velocity spectrum of corresponding interval with described VSP earthquake The seismic velocity of position, it is thus achieved that the rate curve of the described seismic velocity extracted, is carried out the described rate curve obtained Cross the curve obtained after filtering processes as described inverting initial velocity model.

14. VSP pre-drilling pressure forecasting systems as claimed in claim 9, it is characterised in that

Described default inversion algorithm is inversion algorithm based on bayesian theory；

Described corridor stack earthquake record, according to the described inverting initial velocity model set up, is used and presets by described inverting module Inversion algorithm carry out Inversion Calculation, obtain VSP seismic impedance and include:

Simulating the observation data to obtain in advance as dependent variable, the model parameter preset in described inverting initial velocity model is The linear matrix of independent variable；

Utilize first equation first position in described inverting initial velocity model to described corridor stack earthquake record exhibition Open, it is thus achieved that the real seismic record of the described first position relevant to described model parameter；

Using the difference of the described real seismic record of described first position and described corridor stack earthquake record as described sight Survey the disturbance quantity of data；

After described linear matrix is carried out pretreatment, combine with described linear matrix acquisition by the disturbance quantity of described observation data First relational expression；

Described first relational expression is introduced Bayesian formula, it is thus achieved that the Posterior distrbutionp relational expression of described observation data；

The disturbance quantity of model parameter described in each iterative process in Inversion Calculation is obtained according to described Posterior distrbutionp relational expression；

Combine with described inverting initial velocity model acquisition final speed curve by the disturbance quantity of described model parameter；

According to VSP seismic impedance described in described final speed curve and the well density curve acquisition that obtains in advance.

15. VSP pre-drilling pressure forecasting systems as claimed in claim 14, it is characterised in that described first computing module foundation The described VSP seismic impedance obtained calculates the velocity of longitudinal wave of rock and includes: described VSP seismic impedance is substituted into the preset Two algorithms calculate the velocity of longitudinal wave of described rock；

Described method also includes: calculate the vertical of described rock in the second algorithm preset being substituted into by described VSP seismic impedance Before wave velocity, according to the well logging sound wave in the described VSP earthquake record being obtained ahead of time and the density in described well density curve Described second algorithm is fitted by data, simulates the parameter value in described second algorithm；The described parameter value that will simulate Substitute into described second algorithm, and using include the described parameter value simulated the second algorithm as the compressional wave calculating described rock The equation of speed.

16. VSP pre-drilling pressure forecasting systems as claimed in claim 15, it is characterised in that

Described first algorithm is the Fillippone formula improved；

Described second algorithm is Gardner formula；

Described first equation is Taylor's formula.